project-scaffold

Project Scaffold

Initialize and manage research project structure. Creates standardized directories and metadata files that enable seamless handoffs between skills.

Why This Exists

Without scaffolding, every skill asks "Where are your data?" and "Where should I save this?" Project scaffold solves this by:

1. Creating canonical paths that all skills respect
2. Tracking project state so coordinator knows what's done
3. Enabling prerequisite checking ("Can't write findings until analysis is done")
4. Providing audit trail for reproducibility

Project Types

Scaffold supports three project types with different structures:

Type	Use When	Key Directories
qualitative	Interview research, ethnography, content analysis	transcripts, codebook, memos
quantitative	Survey, experiments, secondary data analysis	raw_data, scripts, models, figures
mixed	Combining qual + quant methods	Both sets of directories

Two Modes

Mode 1: New Project

Creates full directory structure and blank metadata files.

/project-scaffold

Title: Why Activists Leave Social Movements
Type: qualitative

Mode 2: Adopt Existing Project

Maps existing directories to canonical paths without moving files.

/project-scaffold adopt

My data is in analysis/
My literature is in lit-review/pdfs/

Workflow

Step 1: Gather Project Info

Ask for:

• Project title (required)
• Project type: qualitative, quantitative, or mixed (required)
• Research question (recommended)
• Mode: new or adopt

Step 1b: Initialize Git

After gathering project info, ensure the project directory is a git repository:

1. If not already a git repo, run git init
2. After creating the directory structure (Step 2), make an initial commit: git add -A && git commit -m "project-scaffold: Initial project structure"

All downstream skills use git for version tracking. Do not create version-suffixed copies of files (e.g., -v2, -final, -working). The git history serves as the version trail.